Proc. Natl. Acad. Sci. USAVol. 86, pp. 2341-2345, April 1989Immunology

Molecular basis for expression of the A48 regulatory idiotope onantibodies encoded by immunoglobulin variable-region genesfrom various families

(nucleotide sequence/hydrophilicity proffle/idiotope-derming region)

HABIB ZAGHOUANI*, F. A. BONILLA*, KATHERYN MEEKt, AND CONSTANTIN BONA*:*Department of Microbiology, The Mount Sinai School of Medicine of the City University of New York, New York, NY 10029; and tDepartment ofMicrobiology, The University of Texas Health Science Center at Dallas, Dallas, TX 75235

Communicated by Edwin D. Kilbourne, December 23, 1988 (received for review July 11, 1988)

ABSTRACT The idiotype defined by the levan-specificBALB/c myeloma protein ABPC48 (A48) has previously beenencountered only in antibodies the variable regions of whichderive from the VHX24 and VK1O gene families. We havedemonstrated expression of the idiotope recognized by themonoclonal anti-A48 idiotype antibody IDA10 on five mono-clonal antibodies from different mouse strains, with differentspecificities including foreign and self antigens and derivingtheir variable regions from families other than VHX24 andVJ1O. We analyzed variable region protein structure (deducedfrom nucleotide sequences) and hydrophilicity profiles ofidiotype' and idiotype- antibodies. We identified four surface-exposed areas (one in the heavy chain and three in the lightchain) that may contribute to expression of the idiotope definedby antibody lDA10.

BALB/c mice immunized with bacterial levan produce anti-bodies against both 82-6 and 82-1 linkages of this polyfruc-tosan (1). A small fraction (<10o) of these antibodies expressthe idiotype found on the BALB/c anti-levan myeloma pro-tein, ABPC48 (A48). However, if these mice are immunizedwith syngeneic polyclonal anti-A48 idiotype antibodies andchallenged with levan, they produce A48 idiotype-bearingimmunoglobulin (2). Some of these immunoglobulins exhibitlevan-binding activity but the vast majority do not (3). It wasspeculated that the latter component represents clones acti-vated through a cross-regulation mediated by anti-idiotype Bor T cells (4). Such clones might have reactivity to selfdeterminants or other unrelated nonselfepitopes. In this studywe screened for A48 idiotype expression 68 monoclonalantibodies (mAbs) originating from various mouse strains andhaving distinct antigen specificities (Table 1).Among this panel, five antibodies were found to bind to the

anti-A48 idiotype mAb IDA10. Because A48 idiotype expres-sion was shown to be restricted to the VHX24 and VK10 genefamilies (12, 13), we studied the variable-region genes en-coding these antibodies. Surprisingly, we found that theirheavy and light chains derived from various gene families.Nucleotide sequence analysis of heavy- and light-chainvariable regions of these idiotype positive (Id') antibodieshas identified amino acid residues that may contribute to theexpression of the A48 idiotype.

MATERIALS AND METHODSThe syngeneic (BALB/c) anti-A48 idiotype mAb IDA10 (14)was provided by Pierre Legrain (Pasteur Institute, Paris). ThemAb 3.14.9, prepared in our laboratory, has been described(12).

Radioimmunoassays. Idiotype expression was assessed byusing a sandwich binding assay performed as follows: micro-titer plates were coated overnight at 4TC with antibodies at 2,ug/ml, washed, and saturated with 1% bovine serum albuminin phosphate-buffered saline. After washing, 90 ng of mAbIDA10 per well was added, and the plates were incubatedovernight at room temperature. Bound mAb IDA10 wasrevealed after 3-hr incubation at 370C with 50,000 cpm of'251-labeled A48 idiotype.For the inhibition assay, microtiter plates were coated

overnight at 4TC with 100 Aul of mAb IDA10 (2 Ag/ml),washed, and saturated with 1% bovine serum albumin inphosphate-buffered saline. The competition for binding tomAb IDA10 was performed by adding serial dilutions ofantibodies together with 20,000 cpm of 125I-labeled mAb3.14.9. After 18-hr incubation at 37°C, the plates were washedand counted.Western (Immunologic) Blot Analysis. The antibodies (1 ,g)

were reduced with mercaptoethanol (25 x 10-2 M), andheavy and light chains were separated by SDS/PAGE (10%acrylamide). Proteins were then transferred to DEAE-membrane by the Bio-Rad Transblot system. After overnightsaturation with 2% (wt/vol) casein in 0.2M sodium carbonatebuffer, the membrane was incubated for 2 hr at roomtemperature in a solution containing 1251I-labeled mAb IDA10or rabbit anti-mouse immunoglobulin antibodies (106 cpm/ml; 10 ml). The membrane was then extensively washed with50 mM Tris HCl, pH 8.2, containing 0.05% Tween 20 andautoradiographed.

Molecular Techniques, Variable Region Gene Probes. Thesource and the size of the VH81X (VH7183 family), theVHNP.B4 (VHJ558 family), the VK8, the VK4, the VK21, and theVK10 probes used in this study have been described in detailelsewhere (11). The bacterial transfection with plasmids,plasmid purification, probe preparation, and [a-32P]dCTPnick-translation were also performed as described (11).Northern (RNA) Blot Analysis. Total RNA was extracted

using the guanidinium thiocyanate procedure (15). Northernblot analysis was performed as described (11).

Cloning. Poly(A)+ mRNA was isolated after one cycle ofselection over a type 3 oligo(dT)-cellulose column accordingto the method described by Aviv and Leder (16). Double-stranded cDNA synthesis was accomplished using a kitmanufactured by Amersham in conjunction with previouslydescribed immunoglobulin C region-specific oligonucleotideprimers (17) commercially prepared by OCS Laboratories(Denton, TX). The double-stranded cDNA was 3' tailed withdCTP and hybridized with a 3' dG-tailed pUC-9 plasmid

Abbreviations: CDR, complementarity-determining region; IDR,idiotope-defining region; mAb, monoclonal antibody; VH, heavy-chain variable region; VK, K chain variable region; Id', idiotypepositive; Id-, idiotype negative.FTo whom reprint requests should be addressed.

2341

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Mar

ch 3

, 202

1

2342 Immunology: Zaghouani et al.

Table 1. VH, origin, and specificity of antibodies examined forA48 idiotype expressionOrigin Antibody (specificity)

BALB/c

mev

129/SVMRL/lprCBA/JDBA/1NZB

BALB/c

MRL/lpr

CBA/JDBA/1SJLmevC57BL/6

BALB/cMRL/lprCBA/J

VH7183HB2(dsDNA), HB12(SA), 62Id(RF), XY10(H3),PY102(H1), 1-15(TG), VM201(H1), B36(Sm),PT109(H1)

S2-6-10(MBP, IF, Thy), S2-17-6(MBP, IF,RBC)

129-48(RF), 129-78(RF)MRL22-46(RF), M13(Sm), M16(Sm)101A1(TG), CP5(Br-RBC)C2(CII)Z51(TG), Z26(Sm), Z121(ds DNA), Z317(dsDNA), Z41(ds DNA)

VHJ558

Y19-10(RF), PY206(H1), PY211(H1), J558*(al-3Dex), MOPC104E*(al-3 Dex)

M56(Sm), Y12(Sm), Y2(Sm), 6B6(Sm),H102(DNA), H130(DNA), H241(DNA),MRL50-8(MBP), 6-19-23(MBP)

8ID2(TG), 8IB1(TG), 10VA2(TG), 84A3(TG)A12(CII), Bll(CII), E8(CII)15-32(MBP)S2-9-2(MBP), S2-10-9(MBP, RBC, Thy)AC5(NP)

VHQ52

HB8(SA), LPS7-4(RF)MRL5-51(RF)B1(TG)

VHS107BALB/c VM202(H1), HPCG15(PCho), HPCG14(PCho),

HPCG8(PCho), HPCG11(PCho),MOPC167*(PCho)

mev UN55-5(NA)VHJ606

BALB/c J606*(inu), W3082*(inu), EPC109*(inu)mev UN42-5(RBC, Thy), S2-14-2(RBC, Thy)

VH36-60BALB/c MOPC460*(TNP), MOPC315*(TNP)DBA/1 E7(CII)The antigen specificity and VH typing of these antibodies are

published elsewhere (5-11). For idiotypic studies these antibodieswere affinity-purified from culture medium on a mAb rat anti-mouseK chain coupled to Sepharose 4B. ds DNA, double-stranded DNA;RF, rheumatoid factor; SA, skin antigen; H1, hemagglutinin of PR8influenza virus; H3, hemagglutinin of X31 influenza virus; TG,thyroglobulin; Sm, Smith antigen; MBP, myelin basic protein; IF,intrinsic factor; Thy, thymocytes; Br-RBC, bromelain-treated eryth-rocytes; CII, collagen type II; al-3Dex, al-3 linked glucose polymer;NP, (4-hydroxy-3-nitrophenyl)acetyl; PCho, phosphorylcholine;NA, nuclear antigen; inu, inulin; TNP, trinitrophenyl.*Myeloma proteins were purified from ascites by chromatographyover protein A or the appropriate antigen coupled to Sepharose 4B.

(Pharmacia) and transfected into JM101 bacteria by pub-lished methods (18).

Sequencing. Two micrograms of supercoiled plasmid DNAwere denatured in 0.2 M sodium hydroxide and hybridized to20 ng of either JP-pUC19 (HindIII) primer (CAGGAAACA-GCTATG) or JP-pUC19 (EcoRI) primer (AAACGACGGC-CAGTG) (prepared and provided by M. Krystal, Mount SinaiSchool of Medicine, New York). The sequencing reactionwas performed using the Sequenase kit manufactured byUnited States Biochemical.

RESULTS

A48 Idiotype Expression on mAbs Having Antigen Specific-ities Other Than Levan. Among 68 mAbs originating from

different mouse strains and exhibiting various antigen-binding specificities, five mAbs were found to express theA48 idiotype as defined by the syngeneic anti-idiotype mAbIDA10. The data depicted in Table 2 show that microtiterwells coated and then incubated with mAb IDA10 bind theId+ 1251I-labeled A48 protein. The binding to Z26 and M56 was2- to 3-fold lower than two prototype A48 Id' antibodies (A48and 3.14.9) and was similar to the binding of mAb IDA10 toanother A48 Id+ levan-binding BALB/c myeloma protein,UPC10 (12). Three other mAbs-XY101, PY102, and Y19-10-showed weaker binding compared with mAbs Z26 andM56 but 2- to 3-fold higher than the binding of IDA10 to mAbJ606, which lacks the A48 idiotype. Expression of the A48idiotype on these antibodies was supported by a competitiveinhibition assay in which these antibodies inhibited thebinding of radiolabeled mAb 3.14.9 to IDA10. As shown inTable 2, mAbs Z26 and M56 strongly inhibited the binding ofmAb 3.14.9 to IDA10, whereas mAbs PY102, XY101, andY19-10 displayed a weaker inhibitory capacity, though stillsignificantly higher compared with mAbs J606 and bovineserum albumin. From these results, we concluded that mAbsZ26 and M56 exhibiting anti-Smith antigen specificity, therheumatoid factor Y19-10, and PY102 and XY101 specific forthe hemagglutinin of influenza viruses expressed an A48idiotope.

Conformational Nature of the A48 Idiotype. Because theA48 idiotype was found on antibodies having distinct antigenactivities encoded by variable regions from different variable-region gene families (Table 2) we studied the contribution ofheavy and light chains to such idiotype expression. TheWestern blotting analysis depicted in Fig. 1 shows that IDA10binds to intact molecules but not to reduced and separatedheavy and light chains. The staining of a duplicate gel with1251I-labeled rabbit anti-mouse immunoglobulin antibody in-stead of IDA10 showed binding to intact antibody moleculesas well as to separated heavy and light chains (data notshown). Thus, it appears that A48 idiotype expression re-quires the contribution of both chains, suggesting its confor-mational nature.

Antibodies Expressing the A48 Idiotype Are Encoded byGenes from Various Families. Northern (RNA) blotting anal-ysis showed that mAb 3.14.9 uses VHX24, whereas mAbsZ26, PY102, and XY101 use heavy-chain variable regiongenes derived from the VH7183 family, mAbs M56 andY19-10 use VH genes from the VHJ558 family. mAb 3.14.9uses VK1O; mAbs M56, Z26, and PY102 use genes derivedfrom the VK8 family, mAb XY101 uses genes from VK21, and

Table 2. Expression of the A48 idiotype by mAbs using differentvariable region gene families

VH gene BindingmAb usage (ref.) assay* Inhibition assayt

ABPC48 X24(19),VK10(13) 8476 ± 275 2,100%o (100)3.14.9 X24(20),VK10j 8184 ± 224 33,100%o (300)UPC10 X24(19),VK10(21) 3220 ± 184 100,95% (2700)Z26 7183t,VK8t 3210 ± 28 200,90% (2700)M56 J558t,VK8t 2228 ± 79 598,85% (2700)PY102 7183*,VK8* 1245 ± 45 >2700,40%o (2700)Y19-10 J558(9),VK4§ 1100 ± 15 >2700,25% (2700)XY101 7183§,V,,21§ 1308 ± 20 >2700,30% (2700)J606 J606(22),VK11(22) 439 ± 61 >8100,1 (8100)BSA 474 ± 54BSA, bovine serum albumin.

*Mean ± SD.tNanograms per well giving 50% inhibition, highest% inhibition seen(highest amount in ng assayed for inhibition).tVariable gene typing was done by Northern blotting and confirmedby nucleotide sequencing.§Variable genes typed by Northern blotting.INo inhibition obtained by indicated amount.

Proc. Natl. Acad. Sci. USA 86 (1989)

Dow

nloa

ded

by g

uest

on

Mar

ch 3

, 202

1

Proc. Natl. Acad. Sci. USA 86 (1989) 2343

44a

chanstoA48 idioe exCp d a b

-VM

6.~6.2Kd

-"W4-4Kd

FIG.I1. Analysis of the contributions of antibody heavy and lightchains to A48 idiotype expression was performed by Westernblotting of unreduced or reduced antibodies detected with radiola-beled IDA10. Antibodies labeled with stars were not reduced, thoselabeled with circles were reduced before electrophoresis. The A48Id' mAbs 3.101.14, 3.9.9, and 3.27.6 (and the myeloma protein J606)are IgG. mAbs Z26, M56, and 3.14.9 (IgM) and A48 (IgA) do not enterthe gel in the unreduced state.

mAb Y19-10 uses genes derived from VK4 (Northern blots arenot shown; the data are summarized in Table 2). Comparisonof variable-region gene nucleotide sequences ofmAbs 3.14.9,Z26, PY102, M56, and Y19-10 to representative germ-linegenes confirmed the assignments to gene families made byNorthern blotting (data not shown).A48 Id' antibodies (including 3.14.9) have previously been

shown to derive their VH regions from a gene of the VHX24gene family (12, 13). This family contains two genes having98% identity in the coding regions (23). VH nucleotidesequences of mAbs Z26, PY102, Y19-10, and M56 werecompared with one of the germ-line genes of the VHX24family (Fig. 2). The mAb Z26 VH sequence has highestidentity with VH441 (75%), whereas mAb Y19-10 has thelowest identity (57%). At the amino acid level, mAb PY102has highest identity (66%), whereas mAbs Y19-10 and M56both have 45% identity (data not shown). These VH regionsuse diversity-region genes from the DSP2 and DFL16 genefamilies and heavy-chain joining genes 1, 3, and 4.

In Fig. 3 we compare the VK nucleotide sequences ofmAbs

Vh441Z26PY 102Y19-10M56

Vh441Z26PY102Y19-10M56

Vh441Z26PY 102Y19-10M56

Vh441Z26PY102Y 19-10M56

3.14.9, PY102, Z26, and M56 with the sequence of an A/Jstrain germ-line VK10 gene. Not surprisingly, the 3-14-9 VK1Ogene has highest identity (99%, 98% at the amino acid level).Of the other genes, M56 has the highest (63%), whereasPY102 has lowest (59%) identity. The amino acid identities ofthese VK regions are 55% and 49o, respectively. All fourantibodies use different JK segments; 3.14.9 uses JK1, PY102uses JK5, Z26 uses JK2, and M56 uses J4.

Correlation of A48 Idiotype Expression to Structural Fea-tures of Both VH and the Variable Region of the K Chain (VK).In an attempt to correlate idiotype expression to variable-region protein structure, we performed the predicted hydro-philicity profiles (26) of the deduced amino acid sequences ofthe heavy and light chains of these A48 Id' and other A48idiotype minus (Id-) mAbs from those listed in Table 1. Sucha profile predicts the surface-exposed amino acids that mightconstitute antigenic determinants. Among the heavy chainsof A48 Id' mAbs, we found only one shared predictedsurface-exposed region containing the aspartic acid in posi-tion 72 (Kabat numbering, ref. 28) that is conserved in mostmurine VH domains (Fig. 4). This region is also hydrophilicin A48 Id- mAbs. However, all six A48 Id' sequences haveeither Asn-Ala or Lys-Ser at positions 73,74, whereas only 2of 11 Id- sequences have one of these combinations.Using a similar analysis for the light chains, we found three

common groups of surface-exposed amino acids in A48 Id'mAbs centered around the conserved Ser-26, the conservedLys-39 and Pro-40, and the conserved Glu-81 and Asp-82(Fig. 4). These regions are also hydrophilic in Id- mAbs. Fiveof five Id' mAbs have either Arg-Ala or Lys-Ser in positions24,25, whereas only one of ten Id- sequences has Arg-Ala inthis location. Five of five Id' light chains have Gly or Glu atposition 42; two of six Id- light chains have Glu at thisposition. Finally, five of five Id' sequences have Glu or Alaat position 80 and Ile or Leu at position 83. Two of six Id-sequences have Ala-80, and one of six has Leu-83. No Id-antibody has all ofthe residues associated with Id' antibodiesin all four surface-exposed areas.

DISCUSSIONStudy of the variable region genes in hybridomas producingantibodies binding polyfructosans and expressing A48 regu-latory idiotopes showed that they derive from the VHX24 andVK10 families (12, 13). Western blotting analysis of thebinding of the anti-A48 idiotype mAb IDA10 to A48 Id'antibodies clearly demonstrated that the idiotope defined bythis anti-idiotype is a conformational determinant requiringboth heavy and light chains. We present data showing that

5 10 15 20 25GAG GTG AAG CTT CTC GAG TCT GGA GGT GGC CTG GTG CAG CCT GGA GGA TCC CTG AAA CTC TCC TGT GCA GCC TCA GGA TTC GAT--A -T- --G --G --- --- --G -AA --- T-A --- A-- --- --- --G --- --- --- --- --- --- --- --- --T --- --- AC---A- --C --G G-G --- --- --G -GA --- T-A --- A-- --- --- --G--- -C- --T --- --- AC-C-- --C C-A --C -AG C-- C-- --G AC- AAA --G --- GG- --- --G -CT --A G-- -GG T-G- --C AGG --T --T --C -A- ACC

C-- --- -C- CC- -AG --G --A A-- --- --G -TC C-A G-- --G A-G- --C AAG --T --T --A -A- --A30 35 40 45 50 52a 55

TTT AGT AGA TAC TGG ATG AGT TGG GTC CGG CAG GCT CCA GGG AAA GGG CTA GAA TGG ATT GGA GAA ATT AAT CCA GAT AGC AGT--C --C --T -AC --- T-- --- --XT-C --- A-- --- -A- --G A-- --G --G -T- G-C -C- -CC --- --- AGT -G- G-T ---

--C --- -AC --T GCC --- TC- --- --T--T --- T-- --G -A- --G A-A --G --G --- G-C -C- TTG --- -C- AGX XX- G-T G-A--C -CC G-C --- -AT --- TAC --G A - -GG --T --A C-G --C --T -GG --- --- --G -GG --- --- --T -GC -AT G-T--C -C- --C --T GTT --- CAC --G AA- --- AAG --T --X C-G X-C --T --G -X- --- --- T-T --- --- --T T-C -AT GAT

60 65 70 75 80 82aACG ATA AAC TAT ACG CCA TCT CTA AAG GAT AAA TTC ATC ATC TCC AGA GAC AAC GCC AAA AAT ACG CTG TAC CTG CAA ATG AGCTAC -CC T-- --- C-A GAC A-- 6-6 --- -GC CG- -- - --- --- --- --- --T --- -GG --C --C --- --- --- --- --- ---

-AC -CC --- --- C-A GAC AG- T-G --- -GC CX- - .- --- --- --- --T--T --- -GG G-C -TC --X --- --- --- --- ---

GGT -CT --- -TC -AT GGG ASG T-C --- GGC -GG GC- -CA C-G A-T GTG --- --A T-- TCC GGC --- GCC --- A-- --- C-C G--GGT -CT --C -AT GAG AAG T-C --A GGC --G GC- -CA C-G A-T TC- --- --A T-- TCC -GC --A GCC --- A-- G-G C-C ---b c 85 90

AAA GTG AGA TCT GAG GAC ACA GCC CTT TAT TAC TGT GCA AGA D REGION-GT C-- -AG --- --- --- --- --- T-G --- --- --- --G --C AAG ACG GGA AAT TAC TAC GGT AGT AGC FL16. 1 Jhl-GT C-- --G --- --- --- -- --- A-G --- --- --- -.---- -TC CG- C-- --- --C FL16.2 Jh4GGC C-- -C- --- -G- --- T-T --G G-C --- --C --- A-- --G CCT --T --C CGG SP2 Jh3-GC C-- -CC --- --- --- T-T --G G-C --- --C--- A-- --G --T --C --- -TG -CC T-G SP2 Jh3

FIG. 2. Comparison ofVH nucleotide sequences ofthe A48 Id' mAbs Z26, PY102, Y19-10, and M56 to the VH441 germ-line gene (23). Dashesindicate nucleotide identity. The diversity-region segment gene families (24) and joining-region genes used are indicated.

Immunology: Zaghouani et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 3

, 202

1

2344 Immunology: Zaghouani et al. Proc. Natl. Acad. Sci. USA 86 (1989)

GAT ATC CAG ATG

-XC --T GT- ---

--C --T GT- ---

--C --T GT- ---

b c d e

CTG XTT AAC AGTCTI TTA AAT AGTCTT TTA AGT AGT

5 10ACA CAG ACT ACA TCC TCC CTG TCT GCC

--- --- T-- C-- CX- --- --- AG- -TG--- --- T-- C-- --- --- --- G-- ATG--- --- T-- C-- --- --- --- G-- ATGf 30 35

GAC ATT AGC AAT TAT TTA AAC TGGTA----- --- -TA --- ---

XGX AGT CAA -AG --C --C --G GC-AGC A-T CAA -AG --C --- -- GC-AGC A-T CAA -AG --C --- --G GC-55 60

ACA TCA AGA TTA CAC TCA GGA GTC CCA----- --AC--- --- ---

GG- -GC GCG ACC XCA CAC --- --- --TG-- --C -CT AGG G-A --T --G --X --TG-- --C -CT AGG G-A --T --G -- --T

80 85GAG CAA GAA GAT ATT GCC ACT TAC TTT

A-- GCT --- --C C-G --A CT- --- -ACC-- GCT --- --CC-G --A XT- --T -AC

TCA AGG TTC AGT_AT

GAT C-C --- -CAGAT C-C --- -TAGAT C-C --- -TA

90TGC CAA CAG GGT

--T --- --A CA---T --G --A CA-

M56 C-- GCT --- --C C-G --A CT- --T -AC --T --G --A CA-

15TCT CTG GGA GAC AGA GTC

--A GCA --- XXX -AG ---

--A G-A --- C-G -AG ---

--A G-A --- C-G -AG ---40

TAT CAG CAG AAA CCA GAT

20 25 27aACC ATC AGT TGC AGG GCA AGT CAG

--T --G --C --T -A- T-C --- --- AGT--T --G -CC --- -A- T-C --- --- AGC--T --G --C --- -A- T-C --- --- AGC

45 50GGA ACT GTT AAA CTC CTG ATC TAC TAC

--C --X A-- --- --- -GG CAG C-- CCX --- X-G T-- --- -X- CGT

--C --- --- --- --- -GA CAG T-- CCT---- --TCA- G-A --- XTT--C --- --- --- --- -GA CAG T-- CC- --- --T C-- G-A --- -TT

65 70 75GGC AGT GGG TCT GGA ACA GAT TAT TCT CTC ACC ATT AGC AAC CTG

--A --- --T --- --G-GX --- -AT A-- --- --A G-C --- -GT G----- --- -- --G --- -TC A-- --T --C --- -GT G----- --- --A- --G --- --- -TC A-- --T -C--- -GT G--

95 95aAAT ACG CTT CCT CGG

_-- --- JklT-- -GX TA- --G JkST-- -GC AC- AXG Jk2T-- -GC AC- --A Jk4

FIG. 3. Comparison of the VK nucleotide sequences of mAbs 3.14.9, PY102, Z26, and M56 to an A/J strain VK1O germ-line gene, VK-GL(25).

this conformational idiotypic determinant can be created bypairing of variable region genes deriving from other variableregion gene families and can be expressed by antibodiesexhibiting various antigen specificities.The groups ofpredicted surface-exposed residues common

to the A48 Id' sequences correspond to some of the IDRsdescribed in the immunoglobulin surface variability analysisof Kieber-Emmons and Kohler (27). The group of homolo-gous residues in the heavy chain corresponds to IDR D,whereas those in the light chain correspond to IDRs A, C, andpart of E (Fig. 4). No heavy or light chains had surface-exposed and IDR-located areas not shared by all A48 Id'mAbs we sequenced. A few surface-exposed areas outsidethe IDRs were unique to one sequence or shared by a few.However, these regions did not appear to correlate withidiotype expression.Of the combinations of residues we have associated with

heavy chains of Id' sequences, Asn-Ala is found in positions

HEAVY CHAIN LIGHT CHAIN

IDR D IDR A IDR C IDR E

61 70 75 22 27 38 45 77 83*4++ *++* ** + +*

A48Id-:ABPC48 pSLKDKFI ISRDNAK3.14.9 pSLKDKFIISRDNAKZ26 DTVKGRfT I SRDNARM56 EKFKGKATLTSDKSSPY102 DSLkgxftISRDNARY19-10 GRFKGRAt 1TVDKSS

A48Id-:HPCG8 aSVKGTf I VSRDTSqHPCG11 aSVKGRfIVSRDTSqHPCG 14 aSVKgrfFVSRDTSqHPCGI5J558 QKfnGLAt 1 TVDKSSMOPC167 aKFKGRfIVSRDTSqCP5 pSIKDRftiFRDNDK3606 ESVKGRfT I SRDDSKW3082 ESVKGRfT I SRDDSKEPC 109 ESVKGRfT I SRDDSKMOPC315 SSLKNRVS I TRDTSE129-48 dTVtgrfTISRDNAKMOPC460

sCRASQsCRASQTCKSSqSCKSSqSCKSSq

scTASescTASLscTASeSCLSSq

SCRSSktcKASQtcqaSAtcRASetcqaSQ

QKPDGTvkQKPDGTvkQKPGQSPKQKPGQSPkKKPGQPXk

qRPGQspqqkPWXs 1kQKPGKApkQKQGKSPgQKPGKAPI

NLEQEDINLEQEDIsVQAEDLsVQAED1SVKAED1

RVKAEDVSLESDDTSLEDEDMsLQPEDFSLEDEDM

SCRSSq qKPGQSPk SREAEDL

FIG. 4. Comparison of common surface-exposed regions of theheavy and light chains of A48 Id' and A48 Id- mAbs. Residuesindicated by uppercase letters have hydrophilicity values (26) greaterthan or equal to 0; lowercase letters denote residues with hydrophi-licity <0. The idiotope-defining regions (IDRs) (27) are indicated. Anasterisk indicates residues conserved in the majority of murineimmunoglobulins. A plus indicates residues that may contribute tothe conformational structure required for expression of the A48idiotope defined by IDA10. All sequences of A48 Id- antibodies arefrom Kabat et al. (28) except for mAbs CP5 (29) and 129-48 (9).

73 and 74 in 18/23 (78%) of VHX24, whereas Lys-Ser is foundin 91/141 (65%) of VHJ558 sequences. Either combination isfound in 15/143 (10%) of sequences from other VH genefamilies (28). The overall high frequency of occurrence ofthese combinations (124/307 or 40%o) would again suggestthat pairing with an appropriate light chain must play animportant role in idiotype expression because of the 68 mAbstested, only five (7%) were found to be Id'.

In the light chains, the residues associated with Id'antibodies in each of the three surface-exposed areas alsocorrelate with variable region gene families in a similar way.However, the concurrence of these combinations of residuesin a single light chain is relatively rare. In a survey of K

light-chain sequences (28), 18/227 (8%) were found to haveall three combinations of residues associated with idiotypeexpression.We assume neither that these residues are all of those that

may contact the anti-idiotype nor that they are themselves allcontacting residues. These amino acids could determineidiotope structure by influencing the spatial location ofresidues that may also occur in Id- mAbs in positionsinappropriate for Id-anti-idiotype interaction. Furthermore,influences of other structures not evident to this analysis mayrender variable regions possessing all four combinations ofresidues Id-. We feel that this type of analysis generatesseveral specific predictions that serve to focus subsequentprobes of variable region structure, such as utilization ofsynthetic peptides or molecular methods such as site-directedmutagenesis.The significance of these findings is 2-fold. (i) Our data

clearly demonstrate that cross-reactive idiotypes are notsolely markers of genes derived from a single variable regiongene family. We have shown that a conformational idiotypicdeterminant of A48, previously thought restricted to variableregions derived from VHX24 and VK10, can be expressed onantibodies deriving their variable regions from various fam-ilies. The ability to form similar variable region structuresusing variable region genes from different families extends tothe combining site, as well. Akolkar et al. (30) have recentlyshown that several antibodies specific for a(1-6) dextranderive their variable regions from various gene families.

(ii) These data shed light on the molecular basis ofidiotype-determined cross-regulatory processes connectingclones specific for foreign or self antigens. The potentialexists for Id' clones to be activated or suppressed byanti-idiotype antibodies or T cells. This possibility should beconsidered when devising strategies for the use of idiotype oranti-idiotype antibodies as therapeutic agents or vaccines.

VK-GL3.14.9PY 102Z26M56

VK-GL3.14.9PY 102Z26M56

VK-GL3.14.9PY102Z26M56

VK-GL3.14.9PY102Z26

Dow

nloa

ded

by g

uest

on

Mar

ch 3

, 202

1

Proc. Natl. Acad. Sci. USA 86 (1989) 2345

This study was supported by Grant IM275 from the AmericanCancer Society, and National Science Foundation Grant DCB-8709711. H.Z. was supported by a grant from the Association de laRecherche sur le Cancer (Villejuif, France). F.A.B. was supportedby a Medical Scientist Training Grant GMO-7280 from the NationalInstitute of General and Medical Sciences.

1. Bona, C., Lieberman, R., Chien, C. C., Mond, J., House, S.,Green, I. & Paul, W. E. (1978) J. Immunol. 120, 1436-1442.

2. Hiernaux, J., Bona, C. & Baker, P. J. (1981) J. Exp. Med. 153,1004-1008.

3. Bona, C. A., Heber-Katz, E. & Paul, W. E. (1981) J. Exp.Med. 153, 951-967.

4. Rubinstein, L. J., Yeh, M. & Bona, C. A. (1982) J. Exp. Med.156, 506-521.

5. Monestier, M., Manheimer-Lory, A., Bellon, B., Painter, C.,Dang, H., Talal, N., Zanetti, M., Schwartz, R., Pisetsky, D.,Kuppers, R., Rose, N., Brochier, J., Klareskog, L., Holmdahl,R., Erlanger, B., Alt, F. & Bona, C. (1986) J. Clin. Invest. 78,753-759.

6. Bellon, B., Manheimer-Lory, A., Monestier, M., Moran, T.,Dimitriu-Bona, A., Alt, F. & Bona, C. (1987) J. Clin. Invest. 79,1044-1053.

7. Holmdahl, R., Nordling, C., Rubin, K., Tarkowski, A. &Klareskog, L. (1986) Scand. J. Immunol. 24, 197-203.

8. Moran, T. M., Monestier, M., Lai, A. C. K., Norton, G.,Reale, M. A., Thompson, M. A., Schulman, J. L., Riblet, R. &Bona, C. A. (1987) Viral Immunol. 1, 1-12.

9. Manheimer-Lory, A. J., Monestier, M., Bellon, B., Alt, F. &Bona, C. A. (1986) Proc. Natl. Acad. Sci. USA 83, 8293-8297.

10. Gearhart, P. J. & Bogenhagen, D. F. (1983) Proc. Natl. Acad.Sci. USA 80, 3439-3443.

11. Painter, C. J., Monestier, M., Chew, A., Dimitriu-Bona, A.,Kashiri, K., Bailey, C., Scott, V. E. & Sidman, C. (1988) J.Exp. Med. 167, 1137-1153.

12. Victor-Kobrin, C., Bonilla, F. A., Bellon, B. & Bona, C. A.(1985) J. Exp. Med. 162, 647-662.

13. Legrain, P. & Buttin, G. (1985) J. Immunol. 134, 3468-3473.

14. Legrain, P., Voegtle, D., Buttin, G. & Cazenave, P. A. (1981)Eur. J. Immunol. 11, 678-685.

15. Chirgwin, J., Przybyla, A., MacDonald, R. & Rutter, W. (1979)Biochemistry 18, 5294-5299.

16. Aviv, H. & Leder, P. (1972) Proc. Nati. Acad. Sci. USA 69,1408-1412.

17. Shlomchik, M., Nemazee, D., Sato, V., Van Snick, J., Carson,D. & Weigert, M. (1986) J. Exp. Med. 164, 407-427.

18. Sikder, S., Kabat, E. A. & Morrison, S. L. (1985) Proc. Nati.Acad. Sci. USA 82, 4045-4049.

19. Auffray, C., Sikorav, J. L., Ollo, R. & Rougeon, F. (1981) Ann.Immunol. (Inst. Pasteur) 132, 77-88.

20. Victor-Kobrin, C., Barak, Z., Bonilla, F. A. & Bona, C. (1987)in Antibodies-Synthesis, Structure, Function, and Immuno-logic Intervention in Disease, eds. Szentivanyi, A., Maurer,P. H. & Janicki, B. W. (Plenum, New York), pp. 91-105.

21. Potter, M., Padlan, E. & Rudikoff, S. (1976) J. Immunol. 117,626-629.

22. Johnson, N., Slankard, J., Paul, L. & Hood, L. (1982) J.Immunol. 128, 302-307.

23. Hartman, A. B. & Rudikoff, S. (1984) EMBO J. 3, 3023-3029.24. Kurosawa, Y. & Tonegawa, S. (1982) J. Exp. Med. 155, 201-

218.25. Meek, K., Sanz, I., Rathbun, G., Nisonoff, A. & Capra, J. D.

(1987) Proc. Natl. Acad. Sci. USA 84, 6244-6248.26. Hopp, T. P. & Woods, K. R. (1981) Proc. Natl. Acad. Sci.

USA 78, 3824-3828.27. Kieber-Emmons, T. & Kohler, H. (1986) Immunol. Rev. 90,

29-48.28. Kabat, E. A., Wu, T. T., Reid-Miller, M., Perry, H. M. &

Gottesman, K. S. (1987) Sequences ofProteins ofImmunolog-ical Interest (Public Health Service, Washington, DC), 4th Ed.

29. Reininger, L., Ollier, P., Poncet, P., Kaushik, A. & Jaton, J.-C.(1987) J. Immunol. 138, 316-323.

30. Akolkar, P. N., Sikder, S. K., Bhattacharya, S. B., Liao, J.,Gruezo, F., Morrison, S. L. & Kabat, E. A. (1987) J. Immunol.138, 4472-4479.

Immunology: Zaghouani et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 3

, 202

1